Binary regression problem with dummy variable(male&female)

[Adam Klaas]

Hello everyone,

below is an example of the data set every variable is binary.

lifestyle_program male age50 age37_49 white medicine reward
0 1 0 1 1 0 0
1 0 0 0 0 0 1
1 1 1 0 1 0 1
1 0 0 1 1 0 1
0 0 1 0 1 0 1
1 0 0 1 0 0 1
0 1 0 1 1 0 1

I want to do a regression for male and for female seperately, however when i try to:

reg medicine lifestyle_program male age50 age37_49 white if male==1

when I run this code I get collinearity, what is the best way to deal with this?

[Carlo Lazzaro]

Adam:
1) if the regressand is a two-level categorical variable, you should go -logit- instead of -regress-;
2) what you experience is frequent when all the precitors are categorical. The only fix is to change your regression specification.

[Mike Lacy]

I'd further note that with -if- restricting the estimation sample to male == 1, the variable -male- will have no variation, and therefore cannot serve as a predictor for both mathematical and conceptual reasons.